#股票交易分享挑战 Kimi at a $50 billion valuation: Bubble or Value?
Opportunities and Hidden Concerns Behind the High Pre-IPO Valuation
When news broke that Moonshot AI’s Kimi was seeking a pre-money valuation of $50 billion in its Series G pre-IPO round, the entire AI venture capital circle erupted. In just eight months, its valuation surged from $4.3 billion to $50 billion, an 11-fold increase. As domestic large-model companies collectively accelerate toward capitalization, this astonishing valuation has become the central topic of industry debate: is it pricing matched by actual strength, or a bubble built up by primary-market capital?
Looking back at the background of this financing round, the Series G round, as the final key private financing round before an IPO, has extremely scarce project allocations.
Market sources indicate that investors have set entry thresholds, requiring institutions to meet asset-management scale requirements, with all funds to be wired by August 15. The strict time constraints and eligibility requirements clearly demonstrate the project’s strong position in financing negotiations and indirectly confirm the market’s enthusiasm for Kimi.
Yet beyond the enthusiasm, debate over whether the valuation is reasonable has never stopped. The first source of confidence supporting the $50 billion valuation is the strength brought by the product and technology.
Since the official launch of the K3 model, Kimi has entered the global first tier in multiple areas, including long-context processing, multimodal understanding, and code generation. This has broken the long-standing perception that domestic large models can only rely on low prices to capture market share.
The product has not only attracted a massive number of users on the consumer side, but its enterprise API business has also raised prices, proving that the model’s capabilities can support higher pricing and allowing it to move away from the old model of relying on subsidies to exchange for usage volume.
Explosive growth in commercialization data is the most important fundamental basis for capital’s willingness to assign a high valuation. According to operating data circulating in the industry, the company’s annualized ARR revenue has risen rapidly, multiplying within just a few months, with API enterprise services contributing more than 70% of revenue.
With consumer memberships and enterprise services advancing on parallel tracks, the company is no longer relying solely on traffic-driven narratives but has genuinely produced a revenue growth curve. Benchmarking the development paths of overseas AI giants, the rapidly growing commercialization prospects have shown the primary market the potential growth of domestic large models. In addition, scarce IPO expectations have brought a significant valuation premium.
Across China, there are only a handful of large-model unicorns that have truly reached the global first tier while also approaching an IPO filing. Amid the global AI boom, a large number of industrial investors and VC institutions want to invest in leading domestic AI assets, but investable targets are limited.
A pre-IPO round means investors are one step closer to an IPO exit after investing, so capital is willing to pay a premium for this scarcity and expected liquidity.
Comparisons with the valuation systems of overseas companies such as OpenAI and Anthropic also show that capital is benchmarking against the global landscape and assigning domestic leaders greater growth potential.
Yet behind the prosperous valuation lie multiple unavoidable real-world challenges, and the $50 billion pricing has also placed a heavy performance burden on the company.
From a valuation-metric perspective, based on the annualized revenue estimates currently circulating in the market, the company’s price-to-sales ratio is extremely high. By comparison, mature companies in the traditional software SaaS industry mostly have price-to-sales ratios in the 10–25 range. Even overseas high-growth AI companies rarely maintain such high multiples over the long term.
A high valuation means the company must sustain explosive revenue growth. To support its current valuation, revenue will need to expand severalfold over the next one to two years. If revenue growth slows, it will be difficult for the existing valuation to obtain fundamental support.
Second, industry competition leaves the company no room to catch its breath. Domestic competitors such as DeepSeek continue to release new versions and seize developers and enterprise customers, while overseas giants are also constantly updating their foundation models, making technological iteration change in an instant. Today’s technological advantage could be caught up with or even surpassed by competitors at any time.
There is no once-and-for-all moat in the large-model sector. The high cost of computing power must also not be overlooked. Even with rapidly rising revenue, large-model training and inference consume massive amounts of high-end chip resources, and companies remain in a stage of substantial losses. As revenue rises, the costs of computing power, research and development, and talent are expanding in tandem.
How to increase gross margins while expanding revenue and move toward profitability is a challenge facing all large-model companies. Attractive revenue figures accompanied by continued losses will remain a key issue for secondary-market investors to scrutinize after the future listing. More importantly, the $50 billion is merely a valuation formed through private transactions in the primary market. The subscription pricing set by a small number of primary-market institutions cannot be equated with the company’s true future market capitalization in the secondary market. Many unicorns that were highly sought after in the primary market have experienced substantial valuation discounts after entering the capital markets. Once the formal IPO takes place, secondary-market investors will scrutinize revenue quality, the scale of losses, and competitive barriers even more rigorously. If commercialization falls short of expectations, the high valuation will face the risk of a correction.
Ultimately, determining whether the $50 billion valuation can hold up comes down to two core tests.
The first test is the ability to deliver on commercialization. Over the next 12 to 18 months, can revenue continue to grow rapidly, will enterprise customer retention remain stable, and can gross margins improve steadily? If revenue continues to rise and gradually absorbs the high price-to-sales ratio, the high valuation will be validated by fundamentals; if growth momentum runs out, the risk of a valuation bubble will emerge.
The second test is the construction of long-term barriers. The company cannot rely solely on foundation-model capabilities; it must also build a second growth curve through agents, industry solutions, and deployment in vertical industries. Relying only on a single API-based model can easily lead to vicious price competition. Only by truly penetrating industries and creating irreplaceable value can it maintain its position in the sector.
Viewed from the perspective of the entire domestic large-model industry, Kimi’s high valuation is not merely about the company itself but is also a microcosm of the sector as a whole. After the hundred-model war, the industry has moved beyond the era of blindly burning money and staking out territory, and capital has begun concentrating on the leaders. The market no longer looks only at parameters and user numbers, but places greater emphasis on commercialization capabilities. Of course, a high valuation is both an honor and a constraint. Capital is willing to bet on the future, but the future ultimately needs real performance to deliver on that bet. Funding can push up a valuation, but a company’s long-term value can only be built up bit by bit through technological iteration and commercial deployment. Once the noise dies down, time will provide the final answer.
#Pre-IPOs第三期KIMI今日开启认购 
